急性缺血性卒中DWI-FLAIR不匹配影像学判读研究进展

doi:10.3969/j.issn.1673-5765.2025.09.016

摘要/Abstract

摘要： 在急性缺血性卒中（acute ischemic stroke，AIS）中，DWI-FLAIR不匹配是指MRI DWI序列上显示有急性梗死病灶（呈现高信号），而对应的FLAIR序列显示该区域呈等信号的一种影像学表现。DWI-FLAIR不匹配反映了梗死核心（不可逆损伤区）与缺血半暗带（潜在可挽救区）在空间分布上的差异。近年来研究表明，基于DWI-FLAIR不匹配的缺血半暗带评估有助于筛选仍存在可挽救脑组织的AIS患者，为发病时间不明（如醒后卒中）的患者是否适宜接受再灌注治疗提供了重要的影像学依据。既往临床研究中，DWI-FLAIR不匹配的判读多依赖医师的视觉评估，因此存在判读异质性问题。随着影像组学与深度学习等技术的发展，DWI-FLAIR不匹配的判读正逐步向精准化与智能化方向迈进。本文系统综述了近年来DWI-FLAIR不匹配影像学判读的进展，并展望了基于精准影像学指导的AIS患者急救体系优化路径。

文章导读： 本文聚焦AIS的DWI-FLAIR不匹配影像学判读，重点阐述了传统视觉判读的局限性，以及人工智能模型的优势，分析了深度学习、放射组学等技术在提升判读精准度上的突破，并对标准化评分系统与智能影像平台的应用前景进行了展望。本文可为临床优化AIS急救体系、改善患者预后，以及推动AIS影像判读向精准化、智能化迈进提供参考。

关键词: 急性缺血性卒中; 弥散加权成像-液体抑制反转恢复不匹配; 影像学

Abstract: In acute ischemic stroke (AIS), DWI-FLAIR mismatch refers to an imaging finding where the MRI DWI sequence shows hyperintensity in the acute infarct core, while the corresponding region on the FLAIR sequence displays isointensity. This finding reflects the spatial distribution discrepancy between the infarct core (irreversibly damaged area) and the ischemic penumbra (potentially salvageable area). Recent studies have demonstrated that assessment based on DWI-FLAIR mismatch can help identify AIS patients who still possess salvageable brain tissue. It provides an important imaging basis for determining the eligibility of patients with unknown onset time (such as wake-up stroke) for reperfusion therapy. In previous clinical studies, the interpretation of DWI-FLAIR mismatch primarily relied on visual assessment by physicians, thus leading to the issue of interpretive heterogeneity. With the development of technologies such as radiomics and deep learning, the interpretation of DWI-FLAIR mismatch is progressively advancing toward precision and intelligence. This article systematically reviews the recent advances in the imaging interpretation of DWI-FLAIR mismatch and prospectively discusses the optimization path of the emergency care system for AIS patients guided by precise imaging.

Key words: Acute ischemic stroke; Diffusion-weighted imaging-fluid attenuated inversion recovery mismatch; Imaging

中图分类号:

李欢, 于向明. 急性缺血性卒中DWI-FLAIR不匹配影像学判读研究进展[J]. 中国卒中杂志, 2025, 20(9): 1198-1202.

LI Huan, YU Xiangming. Research Advances in Imaging Interpretation of DWI-FLAIR Mismatch in Acute Ischemic Stroke[J]. Chinese Journal of Stroke, 2025, 20(9): 1198-1202.

参考文献

[1] GALINOVIC I，DICKEN V，HEITZ J，et al. Homogeneous application of imaging criteria in a multicenter trial supported by investigator training：a report from the WAKE-UP study[J/OL]. Eur J Radiol，2018，104：115-119[2025-07-27]. https://doi.org/10.1016/j.ejrad.2018.05.011.
[2] LEGRAND L，LE BERRE A，SENERS P，et al. FLAIR vascular hyperintensities as a surrogate of collaterals in acute stroke：DWI matters[J]. AJNR Am J Neuroradiol，2023，44（1）：26-32.
[3] PAYABVASH S，TALEB S，BENSON J C，et al. The effects of DWI-infarct lesion volume on DWI-FLAIR mismatch：is there a need for size stratification？[J]. J Neuroimaging，2017，27（4）：392-396.
[4] THOMALLA G，BOUTITIE F，FIEBACH J B，et al. Clinical characteristics of unknown symptom onset stroke patients with and without diffusion-weighted imaging and fluid-attenuated inversion recovery mismatch[J]. Int J Stroke，2018，13（1）：66-73.
[5] CHOI S，NOH D，KIM Y，et al. Magnetic resonance imaging characteristics of ischemic brain infarction over time in a canine stroke model[J]. J Vet Sci，2018，19（1）：137-142.
[6] ZHANG J F，TA N，FU M，et al. Use of DWI-FLAIR mismatch to estimate the onset time in wake-up strokes[J/OL]. Neuropsychiatr Dis Treat，2022，18：355-361[2025-07-27]. https://doi.org/10.2147/NDT.S351943.
[7] WOUTERS A，DUPONT P，CHRISTENSEN S，et al. Association between time from stroke onset and fluid-attenuated inversion recovery lesion intensity is modified by status of collateral circulation[J]. Stroke，2016，47（4）：1018-1022.
[8] CIVRNY J，SEDLACKOVA Z，MALENAK T，et al. Comparison of semi-quantitative and visual assessment of early MRI signal evolution in acute ischaemic stroke[J/OL]. Eur J Radiol Open，2023，10：100488[2025-07-27]. https://doi.org/10.1016/j.ejro.2023.100488.
[9] REGENHARDT R W，BRETZNER M，ZANON ZOTIN M C，et al. Radiomic signature of DWI-FLAIR mismatch in large vessel occlusion stroke[J]. J Neuroimaging，2022，32（1）：63-67.
[10] PFAFF J A，BENDSZUS M，DONNAN G，et al. The impact of the DWI-FLAIR-mismatch in the ECASS-4 trial—a post hoc analysis[J]. Eur Stroke J，2020，5（4）：370-373.
[11] LIU S，XU X Q，CHENG Q G，et al. Simple quantitative measurement based on DWI to objectively judge DWI-FLAIR mismatch in a canine stroke model[J]. Diagn Interv Radiol，2015，21（4）：348-354.
[12] SCHELDEMAN L，WOUTERS A，DUPONT P，et al. Diffusion-weighted imaging and fluid-attenuated inversion recovery quantification to predict diffusion-weighted imaging-fluid-attenuated inversion recovery mismatch status in ischemic stroke with unknown onset[J]. Stroke，2022，53（5）：1665-1673.
[13] CHENG B，BOUTITIE F，NICKEL A，et al. Quantitative signal intensity in fluid-attenuated inversion recovery and treatment effect in the WAKE-UP trial[J]. Stroke，2020，51（1）：209-215.
[14] MADAI V I，GALINOVIC I，GRITTNER U，et al. DWI intensity values predict FLAIR lesions in acute ischemic stroke[J/OL]. PLoS One，2014，9（3）：e92295[2025-07-27]. https://doi.org/10.1371/journal.pone.0092295.
[15] WOUTERS A，CHENG B，CHRISTENSEN S，et al. Automated DWI analysis can identify patients within the thrombolysis time window of 4.5 hours[J/OL]. Neurology，2018，90（18）：e1570-e1577[2025-07-27]. https://doi.org/10.1212/WNL.0000000000005413.
[16] LIU L L，KURGAN L，WU F X，et al. Attention convolutional neural network for accurate segmentation and quantification of lesions in ischemic stroke disease[J/OL]. Med Image Anal，2020，65：101791[2025-07-27]. https://doi.org/10.1016/j.media.2020.101791.
[17] FENG J，ZHANG Q Z，WU F，et al. The value of applying machine learning in predicting the time of symptom onset in stroke patients：systematic review and meta-analysis[J/OL]. J Med Internet Res，2023，25：e44895[2025-07-27]. https://doi.org/10.2196/44895.
[18] ZHU H C，JIANG L，ZHANG H，et al. An automatic machine learning approach for ischemic stroke onset time identification based on DWI and FLAIR imaging[J/OL]. Neuroimage Clin，2021，31：102744[2025-07-27]. https://doi.org/10.1016/j.nicl.2021.102744.
[19] JIANG L，WANG S Y，AI Z P，et al. Development and external validation of a stability machine learning model to identify wake-up stroke onset time from MRI[J]. Eur Radiol，2022，32（6）：3661-3669.
[20] LEE H，LEE E J，HAM S，et al. Machine learning approach to identify stroke within 4.5 hours[J]. Stroke，2020，51（3）：860-866.
[21] POLSON J S，ZHANG H Y，NAEL K，et al. Identifying acute ischemic stroke patients within the thrombolytic treatment window using deep learning[J]. J Neuroimaging，2022，32（6）：1153-1160.
[22] POLSON J，ZHANG H Y，NAEL K，et al. A semi-supervised learning framework to leverage proxy information for stroke MRI analysis[J/OL]. Annu Int Conf IEEE Eng Med Biol Soc，2021：2258-2261[2025-07-27]. https://doi.org/10.1109/EMBC46164.2021.9631098.
[23] DUONG M T，RUDIE J D，WANG J，et al. Convolutional neural network for automated FLAIR lesion segmentation on clinical brain MR imaging[J]. AJNR Am J Neuroradiol，2019，40（8）：1282-1290.
[24] KIM B J，ZHU K，QIU W，et al. Predicting DWI-FLAIR mismatch on NCCT：the role of artificial intelligence in hyperacute decision making[J/OL]. Front Neurol，2023，14：1201223[2025-07-27]. https://doi.org/10.3389/fneur.2023.1201223.